Re: Accord GPS "time to destination" error
 

aljudy wrote:
>
> Hmmm... don't buy that, the only way the navigation system knows where you
> are, and how far you are from where you are going is with satellites.
> Perhaps the Accord navi uses the vehicle speed connection, but that is not
> more accurate than the speed determined from satellites, nor does it help
> when satellites are not available because of obstructions... Al

Navigation systems can be made entirely without GPS (or other satellite
or terrestrial based navigation network). Etak (now TeleAtlas,
http://www.teleatlas.com) did this back in the mid 80's. One version
relied only on speed sensors at the wheels, distance traveled was
calculated based on average of left and right sensor, changes in
direction were calculated based on the difference between left and
right. The path traveled was matched with an electronic map; the system
would assume you were traveling on actual roads and find the best match
of known roads the the measured travel path.

Upon initial installation of the system, a calibration run had to be
made to allow the system to know the characteristics of the wheel
sensors and suspension geometry. An initial absolute location would have
to be entered as well. As tires wore down etc. the system would
recalibrate itself. The system worked well, except for in larger cities
with streets in a regular grid pattern. Here, the system would
occasionally get confused as to what street you were turning onto. E.g.
if you were driving down 2nd ave. for miles, and then made a right turn,
the system would not be accurate enough to know if you were turing on
56th street or 57th street. The system had menu selections for "Skip one
street ahead" and "skip one street back" to overcome this. If you got
wheel spin (e.g. while driving in snow and ice) the system would get
confused as to what direction you were traveling in. On early systems,
the electronic maps were stored on cassette tapes. Later gyroscopes were
added, and the maps were put on CDs.

Gyroscopes were added later (I have one of the early gyroscopes sitting
on my shelf) and experiments were done with using a compass also (turned
out the compass confused the system more than it helped it due to local
variations in the magnetic field etc.).

Navigation without the help of GPS (or similar systems) is called dead
reckoning, and most of the navigation systems that are permanently
installed in cars have some form of dead reckoning in addition to GPS.
If you know where you are (based on GPS data) and then loose the GPS
signal, dead reckoning will give a very accurate estimate of where you
go until you receive a GPS sugnal again.

Re: Accord GPS "time to destination" error
 

aljudy wrote:
>
> Hmmm... don't buy that, the only way the navigation system knows where you
> are, and how far you are from where you are going is with satellites.
> Perhaps the Accord navi uses the vehicle speed connection, but that is not
> more accurate than the speed determined from satellites, nor does it help
> when satellites are not available because of obstructions... Al

Navigation systems can be made entirely without GPS (or other satellite
or terrestrial based navigation network). Etak (now TeleAtlas,
http://www.teleatlas.com) did this back in the mid 80's. One version
relied only on speed sensors at the wheels, distance traveled was
calculated based on average of left and right sensor, changes in
direction were calculated based on the difference between left and
right. The path traveled was matched with an electronic map; the system
would assume you were traveling on actual roads and find the best match
of known roads the the measured travel path.

Upon initial installation of the system, a calibration run had to be
made to allow the system to know the characteristics of the wheel
sensors and suspension geometry. An initial absolute location would have
to be entered as well. As tires wore down etc. the system would
recalibrate itself. The system worked well, except for in larger cities
with streets in a regular grid pattern. Here, the system would
occasionally get confused as to what street you were turning onto. E.g.
if you were driving down 2nd ave. for miles, and then made a right turn,
the system would not be accurate enough to know if you were turing on
56th street or 57th street. The system had menu selections for "Skip one
street ahead" and "skip one street back" to overcome this. If you got
wheel spin (e.g. while driving in snow and ice) the system would get
confused as to what direction you were traveling in. On early systems,
the electronic maps were stored on cassette tapes. Later gyroscopes were
added, and the maps were put on CDs.

Gyroscopes were added later (I have one of the early gyroscopes sitting
on my shelf) and experiments were done with using a compass also (turned
out the compass confused the system more than it helped it due to local
variations in the magnetic field etc.).

Navigation without the help of GPS (or similar systems) is called dead
reckoning, and most of the navigation systems that are permanently
installed in cars have some form of dead reckoning in addition to GPS.
If you know where you are (based on GPS data) and then loose the GPS
signal, dead reckoning will give a very accurate estimate of where you
go until you receive a GPS sugnal again.

Re: Accord GPS "time to destination" error
 

Thanks Randolph for sharing your knowledge here... Al

"Randolph" <trash@junkmail.com> wrote in message
news:4134E9C1.9B892900@junkmail.com...
>
> Navigation systems can be made entirely without GPS (or other satellite
> or terrestrial based navigation network). Etak (now TeleAtlas,
> http://www.teleatlas.com) did this back in the mid 80's. One version
> relied only on speed sensors at the wheels, distance traveled was
> calculated based on average of left and right sensor, changes in
> direction were calculated based on the difference between left and
> right. The path traveled was matched with an electronic map; the system
> would assume you were traveling on actual roads and find the best match
> of known roads the the measured travel path.

Re: Accord GPS "time to destination" error
 

Thanks Randolph for sharing your knowledge here... Al

"Randolph" <trash@junkmail.com> wrote in message
news:4134E9C1.9B892900@junkmail.com...
>
> Navigation systems can be made entirely without GPS (or other satellite
> or terrestrial based navigation network). Etak (now TeleAtlas,
> http://www.teleatlas.com) did this back in the mid 80's. One version
> relied only on speed sensors at the wheels, distance traveled was
> calculated based on average of left and right sensor, changes in
> direction were calculated based on the difference between left and
> right. The path traveled was matched with an electronic map; the system
> would assume you were traveling on actual roads and find the best match
> of known roads the the measured travel path.

Re: Accord GPS "time to destination" error
 

aljudy wrote:
>
> Thanks Randolph for sharing your knowledge here... Al

You are more than welcome. A few years back I had the pleasure of
working with one of the founders of Etak. He was happy to discuss the
details of the navigation system, and also showed me one of the very
early ones that he had installed in his car.

Re: Accord GPS "time to destination" error
 

aljudy wrote:
>
> Thanks Randolph for sharing your knowledge here... Al

You are more than welcome. A few years back I had the pleasure of
working with one of the founders of Etak. He was happy to discuss the
details of the navigation system, and also showed me one of the very
early ones that he had installed in his car.